Society Proceedings / Electroencephalography evidence of systematic errors due to violation of head or source modeling assumptions. In visual studies, qualitatively similar patterns of multiple visual processing areas were observed in each modality. To improve the accuracy of head volume conductor and current source models, we have explored the use of MRI-derived constraints. Component dipole sources were assigned locations and strength based on fMRI and orientations based on anatomical MEL Geometrically complex composite sources estimated from anatomical and functional MRI can then be used to estimate source activation time courses based on MEG. 11. Coexistence of generalized and partial epilepsies. - E.B. Geller, M.E. Lancman, P.C. Van Ness and D.S. Dinner (Department of Neurology, Cleveland Clinic Foundation, Cleveland, OH) Coexisting partial and primary generalized epileptiform discharges are infrequent. We report 6 surgical candidates evaluated at the Cleveland Clinic or UCLA. Mean age of seizure onset was 5.8 years (1.25-24 years). Mean duration of epilepsy was 17.7 years (5-28.5 years). All were neurologically normal and had complex partial seizures recorded. Seizure onset was unilateral temporal in 4, bitemporal in 1, and unilateral frontal in 1. Generalized EEG discharges included 3 Hz spike or polyspike/wave complexes in 5; 3 were asymptomatic, 1 had absence seizures and 2 had myoclonic seizures. MRI or PET showed focal abnormalities in agreement with clinical and EEG data in all. Focal resection was performed in 2, and is planned for 3. We considered secondary bilateral synchrony unlikely since the patients with generalized seizures had classical primary generalized epilepsy syndromes and no patient had the EEG features characteristic of secondary bilateral synchrony. Primary generalized seizures or asymptomatic EEG discharges can exist in patients presenting with partial epilepsies. Such patients may be good epilepsy surgery candidates if symptomatic disabling seizures are proven to be focal after an adequate video EEG evaluation. 12. EEG findings in amygdalar sclerosis. - R.S. McLachlan L.A. Miller (Department of Clinical Neurological Sciences, versity of Western Ontario, London, Ont., Canada)
Isolated amygdalar sclerosis (AS) with a normal hippocampus has been described in a subset of patients who underwent temporal lobectomy (Ann. Neurol., 1993, 33: 622-631). EEG telemetry was reviewed blind to the pathology in 11 patients with AS and 20 with typical mesial temporal sclerosis (MTS). Abnormalities were categorized as unitemporal with or without extratemporal propagation, bitemporal and generalized. Telemetry was done using subdural electrodes in 13 patients, 4 with AS and 9 with MTS. Delta activity and interictal spikes were significantly less likely to be localized to one temporal lobe in AS compared to MTS ( x *, P < 0.05). None of the AS patients had spikes localized to one temporal lobe compared to 45% of MTS patients. Spike frequency was the same in both groups. Seizure onset was less likely to be confined to one temporal lobe in AS (55%) compared to MTS (90%) and seizure propagation was more widespread in AS. Only 3 patients had generalized spike wave, 1 with AS and 2 with MTS. The tendency for electrographic abnormalities in AS to be widely distributed outside the temporal lobe may partially explain why AS patients are less likely to become seizure free than those with MTS following temporal lobectomy. 13. The localization of temporal lobe seizures with scalp EEG: a comparison of sphenoidal and anterior temporal scalp electrodes. - S.V. Pacia and 0. Devinsky @LID/NYU Comprehensive Epilepsy Center, NYU School of Medicine, New York, NY) Several studies comparing sphenoidal electrodes for the identification of mesial reported conflicting results. Moreover, the localize seizures within the temporal lobe
with anterior temporal scalp temporal interictal foci have ability of these electrodes to has not been established. We
and clinical Neurophysiology 95 (199.5) lSP-41P
reviewed 30 seizures recorded with standard 10-20, anterior temporal (T1/2), and sphenoidal electrodes in 12 patients, who subsequently had intracranial EEG localization to the temporal lobe (7 neocortical, 2 mesial, and 1 infero-mesial). Non-invasive EEGs, without patient identification or channel description, were reviewed by an experienced electroencephalographer, in referential (Cz) and longitudinal bipolar montages. The time and channel(s) of ictal onset were noted. Isolated sphenoidal activity on bipolar montage, and high amplitude sphenoidal activity on referential montage, prior to T1/2 changes, were found at ictal onset in both patients with mesial temporal lobe seizures. In contrast, patients with neocortical temporal seizures had simultaneous involvement of temporal chain and anterior temporal electrodes, with or without sphenoidal involvement. The patient with infero-mesial temporal seizures had simultaneous activity in the anterior temporal and sphenoidal electrodes at ictal onset. Sphenoidal electrodes are superior to anterior temporal scalp electrodes (T1/2) in predicting mesial temporal seizure onsets.
14. Four cases of complex partial status with a continuous posterior seizure discharge. - F.W. Sharbrough, T.D. Lagerlund, E.L. So and G.D. Cascino (Section of Electroencephalography, Mayo Clinic, Rochester, MN)
Complex partial status with a continuous seizure discharge lasting more than 30 min is an unusual event. Cases have been reported with either frontal or posterior seizure discharges. In the past year on our Epilepsy Monitoring Unit, we have observed 4 cases of status with a seizure discharge lasting longer than 30 min and in which, at least in the latter stages of al1 seizures, there was a prolonged continuous discharge in the posterior head region. One of these originated from a post-surgical occipital focus. This case has become seizure free after a left occipital lobectomy. In the other 3 patients, the seizure was initiated from an anterior temporal focus. Two of these latter 3 cases had hippocampal atrophy and are now seizure free after a left anterior temporal lobectomy. The spatial evolution of these seizure discharges will be demonstrated with Laplacian montage and the following question explored: Is there something intrinsic about the posterior neuronal organization that allows this region to more readily support a prolonged continuous seizure discharge even in cases where the seizure begins in a lower threshold anterior medial temporal focus?
15. Comparison of two methods for simulating cortical surface potential fields. - J.S. Semel, L. Ke and RD. Sidman (Department of Mathematics, University of Southwestern Louisiana, Lafayette, LA)
Two methods of simulating cortical surface potential maps are compared. These methods are the spherical harmonic expansion method (SHE: Lagerlund et al., 1993) and the cortical imaging technique (CIT: Sidman, 19911. Artificial sources are used to generate scalp-recorded potentials. These data are then used as input to SHE and CIT to produce theoretical cortical surface voltages which are then compared pointwise. In SHE, the head is simulated by a layered spherical medium where the layers represent the brain, skull, and scalp. The potential can be expanded in a series of spherical harmonics where the coefficients are determined by boundary conditions at the surface of each layer. A truncation of this series is then used to approximate the potential at the surface of the “brain.” In CIT, the head is simulated by a homogeneous conducting sphere in which the potential can be written in closed form. This function can be modified to account for the skull and scalp layers (At-y et al., 1981). After comparing the accuracy of these methods we will also discuss their relative computational speeds, which will eventually be a factor in analyzing large datasets (e.g., a 22 megabyte, 5 min EEG epoch containing epileptiform spikes).